Mail Thread

Mail Thread — Read a Full Email Conversation

Thread: $ARGUMENTS

Step 1: Find the thread

Option A: Search by subject keyword

DB="$HOME/Library/Mail/V10/MailData/Envelope Index"
QUERY=""

sqlite3 "$DB" "
SELECT m.ROWID, m.conversation_id, s.subject,
       datetime(m.date_received,'unixepoch','localtime') as dt,
       a.address as sender
FROM messages m
JOIN subjects  s ON m.subject = s.ROWID
JOIN addresses a ON m.sender  = a.ROWID
WHERE s.subject LIKE '%${QUERY}%' AND m.deleted=0
ORDER BY m.date_received DESC LIMIT 10;" 2>/dev/null

Option B: If ROWID provided directly

sqlite3 "$DB" "
SELECT conversation_id FROM messages WHERE ROWID = ;" 2>/dev/null

Step 2: Get all messages in the thread by conversation_id

CONV_ID=

sqlite3 "$DB" "
SELECT m.ROWID,
       datetime(m.date_received,'unixepoch','localtime') as dt,
       a.address as sender, a.comment as name,
       s.subject, mb.url as mailbox, m.read
FROM messages m
JOIN subjects  s  ON m.subject = s.ROWID
JOIN addresses a  ON m.sender  = a.ROWID
JOIN mailboxes mb ON m.mailbox = mb.ROWID
WHERE m.conversation_id = ${CONV_ID}
  AND m.deleted = 0
ORDER BY m.date_received ASC;" 2>/dev/null

Step 3: Read all message bodies

python3 ~/.claude/skills/_mail-shared/parser.py    ...

Step 4: Reconstruct the conversation

Present chronologically:

• Who sent each message and when
• Key content of each message (summarize long ones, quote short ones)
• Reply chain structure

Step 5: Summarize the thread

After showing the conversation:

• What was the original question/issue?
• What was discussed/debated?
• What was decided or concluded?
• Are there any open questions or pending actions?
• Who is waiting on whom?

Output Format

Thread: [Subject] [N messages | started: date | participants: list]

---

[Date] [Sender Name] [→ Reply to: ...] > [message content, summarized if long]

---

[repeat for each message]

---

Summary: [2-3 sentence summary of the thread outcome]

Status: [Resolved / Ongoing / Waiting for response] Next action: [if any]

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: aashari
• Source: aashari/ai-agent-skills
• License: MIT

Install and usage instructions live in the source repository linked above.